Weakly Supervised Contrastive Learning for Chest X-Ray Report Generation
This work addresses the challenge of generating informative radiology reports for clinical diagnosis, which is incremental as it builds on existing encoder-decoder models with a novel loss function.
The paper tackled the problem of generating clinically accurate radiology reports from chest X-rays by addressing dataset bias where normal findings dominate, using a weakly supervised contrastive loss to improve text outputs. The method outperformed previous work on clinical correctness and text generation metrics for two public benchmarks.
Radiology report generation aims at generating descriptive text from radiology images automatically, which may present an opportunity to improve radiology reporting and interpretation. A typical setting consists of training encoder-decoder models on image-report pairs with a cross entropy loss, which struggles to generate informative sentences for clinical diagnoses since normal findings dominate the datasets. To tackle this challenge and encourage more clinically-accurate text outputs, we propose a novel weakly supervised contrastive loss for medical report generation. Experimental results demonstrate that our method benefits from contrasting target reports with incorrect but semantically-close ones. It outperforms previous work on both clinical correctness and text generation metrics for two public benchmarks.